Search Results (5,187)

In distance learning environments, learner engagement directly impacts attention, motivation, and academic performance. Signs of fatigue, negative affect, or critical remarks can warn of growing disengagement and potential dropout. However, most existing approaches rely on a single modality, visual or text-based, without providing a general view of learners’ cognitive and affective states. We propose a multimodal system that integrates three complementary analyzes: (1) a CNN-LSTM model augmented with warning signs such as PERCLOS and yawning frequency for fatigue detection, (2) facial emotion recognition by EmoNet and an LSTM to handle temporal dynamics, and (3) sentiment analysis of feedback by a fine-tuned BERT model. It was evaluated on three public benchmarks: DAiSEE for fatigue, AffectNet for emotion, and MOOC Review (Coursera) for sentiment analysis. The results show a precision of 88.5% for fatigue detection, 70% for emotion detection, and 91.5% for sentiment analysis. Aggregating these cues enables an accurate identification of disengagement periods and triggers individualized pedagogical interventions. These results, although based on independently sourced datasets, demonstrate the feasibility of an integrated approach to detecting disengagement and open the door to emotionally intelligent learning systems with potential for future work in real-time content personalization and adaptive learning assistance. Full article

The parameter identification of Autonomous Underwater Vehicles (AUVs) serves as a fundamental basis for achieving high-precision motion control, state monitoring, and system development. Currently, AUV parameter identification typically relies on the complete motion information obtained from onboard sensors. However, in practical applications, it is often challenging to accurately measure key state variables such as velocity and angular velocity, resulting in incomplete measurement data that compromises identification accuracy and model reliability. This issue is particularly pronounced in vertical motion tasks involving low-speed, large pitch angles, and highly maneuverable conditions, where the strong coupling and nonlinear characteristics of underwater vehicles become more significant. Traditional hydrodynamic models based on full-state measurements often suffer from limited descriptive capability and difficulties in parameter estimation under such conditions. To address these challenges, this study investigates a parameter identification method for AUVs operating under vertical, large-amplitude maneuvers with constrained measurement information. A control autoregressive (CAR) model-based identification approach is derived, which requires only pitch angle, vertical velocity, and vertical position data, thereby reducing the dependence on complete state observations. To overcome the limitations of the conventional Recursive Least Squares (RLS) algorithm—namely, its slow convergence and low accuracy under rapidly changing conditions—a Multi-Innovation Least Squares (MILS) algorithm is proposed to enable the efficient estimation of nonlinear hydrodynamic characteristics in complex dynamic environments. The simulation and experimental results validate the effectiveness of the proposed method, demonstrating high identification accuracy and robustness in scenarios involving large pitch angles and rapid maneuvering. The results confirm that the combined use of the CAR model and MILS algorithm significantly enhances model adaptability and accuracy, providing a solid data foundation and theoretical support for the design of AUV control systems in complex operational environments. Full article

To address the inefficiency of traditional concrete crack detection methods and the heavy reliance of supervised learning on extensive labeled data, in this study, an intelligent assessment method of concrete damage based on pseudo-label semi-supervised learning and fractal geometry theory is proposed to solve two core tasks: one is binary classification of pixel-level cracks, and the other is multi-category assessment of damage state based on crack morphology. Using three-channel RGB images as input, a dual-path collaborative training framework based on U-Net encoder–decoder architecture is constructed, and a binary segmentation mask of the same size is output to achieve the accurate segmentation of cracks at the pixel level. By constructing a dual-path collaborative training framework and employing a dynamic pseudo-label refinement mechanism, the model achieves an F1-score of 0.883 using only 50% labeled data—a mere 1.3% decrease compared to the fully supervised benchmark DeepCrack (F1 = 0.896)—while reducing manual annotation costs by over 60%. Furthermore, a quantitative correlation model between crack fractal characteristics and structural damage severity is established by combining a U-Net segmentation network with the differential box-counting algorithm. The experimental results demonstrate that under a cyclic loading of 147.6–221.4 kN, the fractal dimension monotonically increases from 1.073 (moderate damage) to 1.189 (failure), with 100% accuracy in damage state identification, closely aligning with the degradation trend of macroscopic mechanical properties. In complex crack scenarios, the model attains a recall rate (Re = 0.882), surpassing U-Net by 13.9%, with significantly enhanced edge reconstruction precision. Compared with the mainstream models, this method effectively alleviates the problem of data annotation dependence through a semi-supervised strategy while maintaining high accuracy. It provides an efficient structural health monitoring solution for engineering practice, which is of great value to promote the application of intelligent detection technology in infrastructure operation and maintenance. Full article

Rising volumes of textile waste necessitate the development of more efficient recycling systems, with a primary focus on the optimization of sorting technologies. Near-infrared (NIR) spectroscopy is a state-of-the-art method for fiber identification; however, its accuracy in quantifying textile blends, particularly common polyester/cotton blend textiles, still requires refinement. This study explores the potential and limitations of NIR spectroscopy for quantifying cotton content in post-consumer textiles. A lab-scale NIR sorter and a handheld NIR spectrometer in complementary wavelength ranges were applied to a diverse range of post-consumer textile samples to test model accuracies. Results show that the commonly assumed 10% accuracy threshold in industrial sorting can be exceeded, especially when excluding textiles with <35% cotton content. Identifying and excluding the range of non-linearity significantly improved the model’s performance. The final models achieved an RMSEP of 6.6% and bias of −0.9% for the NIR sorter and an RMSEP of 3.1% and bias of −0.6% for the handheld NIR spectrometer. This study also assessed how textile characteristics—such as color, structure, product type, and alkaline treatment—affect spectral behavior and model accuracy, highlighting their importance for refining quantification when high-purity inputs are needed. By identifying current limitations and potential sources of errors, this study provides a foundation for improving NIR-based models. Full article

Multimodal object detection leverages the complementary characteristics of visible (RGB) and infrared (IR) imagery, making it well-suited for challenging scenarios such as low illumination, occlusion, and complex backgrounds. However, most existing fusion-based methods rely on static or heuristic strategies, limiting their adaptability to dynamic environments. To address this limitation, we propose CLSANet, a cognitive learning-based self-adaptive network that enhances detection performance by dynamically selecting and integrating modality-specific features. CLSANet consists of three key modules: (1) a Dominant Modality Identification Module that selects the most informative modality based on global scene analysis; (2) a Modality Enhancement Module that disentangles and strengthens shared and modality-specific representations; and (3) a Self-Adaptive Fusion Module that adjusts fusion weights spatially according to local scene complexity. Compared to existing methods, CLSANet achieves state-of-the-art detection performance with significantly fewer parameters and lower computational cost. Ablation studies further demonstrate the individual effectiveness of each module under different environmental conditions, particularly in low-light and occluded scenes. CLSANet offers a compact, interpretable, and practical solution for multimodal object detection in resource-constrained settings. Full article

In the era of smart agriculture, accurate identification of plant species is critical for effective crop management, biodiversity monitoring, and the sustainable use of medicinal resources. However, existing deep learning approaches often underperform when applied to fine-grained plant classification tasks due to the lack of domain-specific, high-quality datasets and the limited representational capacity of traditional architectures. This study addresses these challenges by introducing a novel, well-curated leaf image dataset consisting of 39 classes of medicinal and aromatic plants collected from the Aromatic and Medicinal Plant Research Station in Odakkali, Kerala, India. To overcome performance bottlenecks observed with a baseline Convolutional Neural Network (CNN) that achieved only 44.94% accuracy, we progressively enhanced model performance through a series of architectural innovations. These included the use of a pre-trained VGG16 network, data augmentation techniques, and fine-tuning of deeper convolutional layers, followed by the integration of Squeeze-and-Excitation (SE) attention blocks. Ultimately, we propose a hybrid deep learning architecture that combines VGG16 with Batch Normalization, Gated Recurrent Units (GRUs), Transformer modules, and Dilated Convolutions. This final model achieved a peak validation accuracy of 95.24%, significantly outperforming several baseline models, such as custom CNN (44.94%), VGG-19 (59.49%), VGG-16 before augmentation (71.52%), Xception (85.44%), Inception v3 (87.97%), VGG-16 after data augumentation (89.24%), VGG-16 after fine-tuning (90.51%), MobileNetV2 (93.67), and VGG16 with SE block (94.94%). These results demonstrate superior capability in capturing both local textures and global morphological features. The proposed solution not only advances the state of the art in plant classification but also contributes a valuable dataset to the research community. Its real-world applicability spans field-based plant identification, biodiversity conservation, and precision agriculture, offering a scalable tool for automated plant recognition in complex ecological and agricultural environments. Full article

The growth and sustainability of today’s global economy heavily relies on smooth maritime operations. The increasing security concerns to marine environments pose complex security challenges, such as smuggling, illegal fishing, human trafficking, and environmental threats, for traditional surveillance methods due to their limitations. Artificial intelligence (AI), particularly deep learning, has offered strong capabilities for automating object detection, anomaly identification, and situational awareness in maritime environments. In this paper, we have reviewed the state-of-the-art deep learning models mainly proposed in recent literature (2020–2025), including convolutional neural networks, recurrent neural networks, Transformers, and multimodal fusion architectures. We have highlighted their success in processing diverse data sources such as satellite imagery, AIS, SAR, radar, and sensor inputs from UxVs. Additionally, multimodal data fusion techniques enhance robustness by integrating complementary data, yielding more detection accuracy. There still exist challenges in detecting small or occluded objects, handling cluttered scenes, and interpreting unusual vessel behaviours, especially under adverse sea conditions. Additionally, explainability and real-time deployment of AI models in operational settings are open research areas. Overall, the review of existing maritime literature suggests that deep learning is rapidly transforming maritime domain awareness and response, with significant potential to improve global maritime security and operational efficiency. We have also provided key datasets for deep learning models in the maritime security domain. Full article

The Junggar Basin contains a large amount of coal resources and is an important coal production base in China. The coal seam in Zhundong coalfield has a large single-layer thickness and high content of inertinite, but large particle Fe-sulphide minerals are associated with coal seams in some mining areas. A series of economic and environmental problems caused by the combustion of large-grained Fe-sulphide minerals in coal have seriously affected the economic, clean and efficient utilization of coal. In this paper, the ultra-thick coal seam of the Xishanyao formation in the Yihua open-pit mine of the Zhundong coalfield is taken as the research object. Through the analysis of coal quality, X-ray fluorescence spectrometer test of major elements in coal, inductively coupled plasma mass spectrometry test of trace elements, SEM-Raman identification of Fe-sulphide minerals in coal and LA-MC-ICP-MS test of sulfur isotope of marcasite, the coal quality characteristics, main and trace element characteristics, macro and micro occurrence characteristics of Fe-sulphide minerals and sulfur isotope characteristics of marcasite in the ultra-thick coal seam of the Xishanyao formation are tested. On this basis, the occurrence state and genesis of large particle Fe-sulphide minerals in the ultra-thick coal seam of the Xishanyao formation are clarified. The main results and understandings are as follows: (1) the occurrence state of Fe-sulphide minerals in extremely thick coal seams is clarified. The Fe-sulphide minerals in the extremely thick coal seam are mainly marcasite, and concentrated in the YH-2, YH-3, YH-8, YH-9, YH-14, YH-15 and YH-16 horizons. Macroscopically, Fe-sulphide minerals mainly occur in three forms: thin film Fe-sulphide minerals, nodular Fe-sulphide minerals, and disseminated Fe-sulphide minerals. Microscopically, they mainly occur in four forms: flake, block, spearhead, and crack filling. (2) The difference in sulfur isotope of marcasite was discussed, and the formation period of marcasite was preliminarily divided. The overall variation range of the δ³⁴S value of marcasite is wide, and the extreme values are quite different. The polyflake marcasite was formed in the early stage of diagenesis and the δ³⁴S value was negative, while the fissure filling marcasite was formed in the late stage of diagenesis and the δ³⁴S value was positive. (3) The coal quality characteristics of the thick coal seam were analyzed. The organic components in the thick coal seam are mainly inertinite, and the inorganic components are mainly clay minerals and marcasite. (4) The difference between the element content in the thick coal seam of the Zhundong coalfield and the average element content of Chinese coal was compared. The major element oxides in the thick coal seam are mainly CaO and MgO, followed by SiO₂, Al₂O₃, Fe₂O₃ and Na₂O. Li, Ga, Ba, U and Th are enriched in trace elements. (5) The coal-accumulating environment characteristics of the extremely thick coal seam are revealed. The whole thick coal seam is formed in an acidic oxidation environment, and the horizon with Fe-sulphide minerals is in an acidic reduction environment. The acidic reduction environment is conducive to the formation of marcasite and is not conducive to the formation of pyrite. (6) There are many matrix vitrinite, inertinite content, clay content, and terrigenous debris in the extremely thick coal seam. The good supply of peat swamp, suitable reduction environment and pH value, as well as groundwater leaching and infiltration, together cause the occurrence of large-grained Fe-sulphide minerals in the extremely thick coal seam of the Xishanyao formation in the Zhundong coalfield. Full article

Vehicle re-identification (Re-ID) is a critical computer vision task that aims to match the same vehicle across spatially distributed cameras, especially in the context of remote sensing imagery. While prior research has primarily focused on Re-ID using remote sensing images captured from similar, typically elevated viewpoints, these settings do not fully reflect complex aerial-ground collaborative remote sensing scenarios. In this work, we introduce a novel and challenging task: aerial-ground cross-view vehicle Re-ID, which involves retrieving vehicles in ground-view image galleries using query images captured from aerial (top-down) perspectives. This task is increasingly relevant due to the integration of drone-based surveillance and ground-level monitoring in multi-source remote sensing systems, yet it poses substantial challenges due to significant appearance variations between aerial and ground views. To support this task, we present AGID (Aerial-Ground Vehicle Re-Identification), the first benchmark dataset specifically designed for aerial-ground cross-view vehicle Re-ID. AGID comprises 20,785 remote sensing images of 834 vehicle identities, collected using drones and fixed ground cameras. We further propose a novel method, Enhanced Self-Correlation Feature Computation (ESFC), which enhances spatial relationships between semantically similar regions and incorporates shape information to improve feature discrimination. Extensive experiments on the AGID dataset and three widely used vehicle Re-ID benchmarks validate the effectiveness of our method, which achieves a Rank-1 accuracy of 69.0% on AGID, surpassing state-of-the-art approaches by 2.1%. Full article

Radio Frequency Identification (RFID) technology is widely applied in various scenarios, including logistics tracking, supply chain management, and target monitoring. In these contexts, the malicious cloning of legitimate tag information can lead to sensitive data leakage and disrupt the normal acquisition of tag information by readers, thereby threatening personal privacy and corporate security and incurring significant economic losses. Although some efforts have been made to detect cloning attacks, the presence of missing tags in RFID systems can obscure cloned ones, resulting in a significant reduction in identification efficiency and accuracy. To address these problems, we propose the block-based cloned tag identification (BCTI) protocol for identifying cloning attacks in the presence of missing tags. First, we introduce a block indicator to sort all tags systematically and design a block mechanism that enables tags to respond repeatedly within a block with minimal time overhead. Then, we design a superposition strategy to further reduce the number of verification times, thereby decreasing the execution overhead. Through an in-depth analysis of potential tag response patterns, we develop a precise method to identify cloning attacks and mitigate interference from missing tags in probabilistic cloning attack scenarios. Moreover, we perform parameter optimization of the BCTI protocol and validate its performance across diverse operational scenarios. Extensive simulation results demonstrate that the BCTI protocol meets the required identification reliability threshold and achieves an average improvement of $24.01\%$ in identification efficiency compared to state-of-the-art solutions. Full article

Background: Automated surgical navigation can be separated into three stages: (1) organ identification and localization, (2) identification of the organs requiring further surgery, and (3) automated planning of the operation path and steps. With its ideal visual and operating system, the da Vinci surgical system provides a promising platform for automated surgical navigation. This study focuses on the first step in automated surgical navigation by identifying organs in gynecological surgery. Methods: Due to the difficulty of collecting da Vinci gynecological endoscopy data, we propose DeepVinci, a novel end-to-end high-performance encoder–decoder network based on convolutional neural networks (CNNs) for pixel-level organ semantic segmentation. Specifically, to overcome the drawback of a limited field of view, we incorporate a densely multi-scale pyramid module and feature fusion module, which can also enhance the global context information. In addition, the system integrates an edge supervision network to refine the segmented results on the decoding side. Results: Experimental results show that DeepVinci can achieve state-of-the-art accuracy, obtaining dice similarity coefficient and mean pixel accuracy values of 0.684 and 0.700, respectively. Conclusions: The proposed DeepVinci network presents a practical and competitive semantic segmentation solution for da Vinci gynecological surgery. Full article

Background/Objectives: Metabolic dysfunction-associated steatotic liver disease (MASLD) represents a major global health challenge characterized by complex adipose–liver interactions mediated by adipokines and hepatokines. Despite rapid field evolution, a comprehensive understanding of research trends and translational advances remains fragmented. This study systematically maps the scientific landscape through bibliometric analysis, identifying emerging domains and future clinical translation directions. Methods: A comprehensive bibliometric analysis of 1002 publications from 2004 to 2025 was performed using thematic mapping, temporal trend evaluation, and network analysis. Analysis included geographical and institutional distributions, thematic cluster identification, and research paradigm evolution assessment, focusing specifically on adipokine–hepatokine signaling mechanisms and clinical implications. Results: The United States and China are at the forefront of research output, whereas European institutions significantly contribute to mechanistic discoveries. The thematic map analysis reveals the motor/basic themes residing at the heart of the field, such as insulin resistance, fatty liver, metabolic syndrome, steatosis, fetuin-A, and other related factors that drive innovation. Basic clusters include metabolic foundations (obesity, adipose tissue, FGF21) and adipokine-centered subjects (adiponectin, leptin, NASH). New themes focus on inflammation, oxidative stress, gut microbiota, lipid metabolism, and hepatic stellate cells. Niche areas show targeted fronts such as exercise therapies, pediatric/novel adipokines (chemerin, vaspin, omentin-1), and advanced molecular processes that focus on AMPK and endoplasmic-reticulum stress. Temporal analysis shows a shift from single liver studies to whole models that include the gut microbiota, mitochondrial dysfunction, and interactions between other metabolic systems. The network analysis identifies nine major clusters: cardiovascular–metabolic links, adipokine–inflammatory pathways, hepatokine control, and new therapeutic domains such as microbiome interventions and cellular stress responses. Conclusions: In summary, this study delineates current trends and emerging areas within the field and elucidates connections between mechanistic research and clinical translation to provide guidance for future research and development in this rapidly evolving area. Full article

This paper addresses the problem of loss of control during flight caused by failures of flight control surfaces. It presents a study of an emergency thrust control system based on linear-quadratic control with integral action. The research encompasses an analysis of thrust modulation control characteristics, a review of existing control systems, and a detailed description of the development process, including the research platform configuration, identification of the aircraft state-space model, control law design, integration of system components within the MATLAB and Simulink environment, and software-in-the-loop testing conducted in the X-Plane 11 flight simulator using a Boeing 757-200 model. The study also investigates the issue of control channel cross-coupling and its impact on simultaneous control of the aircraft’s longitudinal and lateral dynamics. The simulation results demonstrate that the proposed emergency system provides adequate controllability, with settling times of approximately 12 s for achieving a flight path angle setpoint of +5°, and 13 s for attaining a maximum (limited) roll angle of 20°, achieved in separate manoeuvres. Furthermore, simulated landing attempts suggest that the system could potentially enable successful landings at approach speeds significantly higher than standard recommendations. However, further investigation is required to address decoupling of control channels, ensure system stability, and evaluate control performance across a broader range of aircraft configurations. Full article

To address the issue of low accuracy in identifying the transition states of rolling bearing performance degradation when relying solely on vibration signals, this study proposed a vibration–temperature fusion-based adaptive method for bearing performance degradation assessments. First, a multidimensional time–frequency feature set was constructed by integrating vibration acceleration and temperature signals. Second, a novel composite sensitivity index (CSI) was introduced, incorporating the trend persistence, monotonicity, and signal complexity to perform preliminary feature screening. Mutual information clustering and regularized entropy weight optimization were then combined to reselect highly sensitive parameters from the initially screened features. Subsequently, an adaptive feature fusion method based on auto-associative kernel regression (AFF-AAKR) was introduced to compress the data in the spatial dimension while enhancing the degradation trend characterization capability of the health indicator (HI) through a temporal residual analysis. Furthermore, the entropy weight method was employed to quantify the information entropy differences between the vibration and temperature signals, enabling dynamic weight allocation to construct a comprehensive HI. Finally, a dual-criteria adaptive bottom-up merging algorithm (DC-ABUM) was proposed, which achieves bearing life-stage identification through error threshold constraints and the adaptive optimization of segmentation quantities. The experimental results demonstrated that the proposed method outperformed traditional vibration-based life-stage identification approaches. Full article

To address the wear life prediction challenge of Guide Cones in passive compliant connectors under dynamic loads within specialized equipment, this study proposes a dynamic wear modeling and life assessment method based on the improved Archard model. Through integrated theoretical modeling, finite element simulation, and experimental validation, we establish a bidirectional coupling framework analyzing dynamic contact mechanics and wear evolution. By developing phased contact state identification criteria and geometric constraints, a transient load calculation model is established, revealing dynamic load characteristics with peak contact forces reaching 206.34 N. A dynamic contact stress integration algorithm is proposed by combining Archard’s theory with ABAQUS finite element simulation and ALE adaptive meshing technology, enabling real-time iterative updates of wear morphology and contact stress. This approach constructs an exponential model correlating cumulative wear depth with docking cycles (R² = 0.997). Prototype experiments demonstrate a mean absolute percentage error (MAPE) of 14.6% between simulated and measured wear depths, confirming model validity. With a critical wear threshold of 0.8 mm, the predicted service life reaches 45,270 cycles, meeting 50-year operational requirements (safety margin: 50.9%). This research provides theoretical frameworks and engineering guidelines for wear-resistant design, material selection, and life evaluation in high-reliability automatic docking systems. Full article